GLACIEBS OF UNAKWIK INLET AND COLLEGE FIORD 315 On the western side of lower College Fiord there is continuous thick, mature, coniferous forest at sea level as far north as Barnard Glacier; but beyond that there are only scattered groves, interrupted by the barren zones of the cascading glaciers. The spruces surrounding the barren bulb of Vassar Glacier grow only near sea level, not extending up the steepened fiord walls as they do between Barnard and Wellesley Glaciers. One of the spruces at sea level, near the north side of Vassar Glacier, was 16 inches in diameter five feet above the ground and must have grown a century old or more. Between Vassar and Bryn Mawr Glaciers there are only a few spruces at sea level; but at about the hanging valley level there are two large groves. Between Bryn Mawr and Smith Glaciers there are scattered mature spruces at the hanging valley level near the former, at about the 1000-foot level near the latter, and at sea level on the south side of Smith Glacier. Between Smith and Harvard Glaciers there are scattered spruces nearly down to sea level, extending much closer to Harvard Glacier than Gilbert thought; and one which had just been pushed over by the advancing glacier in 1910 was over a foot in diameter and therefore probably nearly or quite a century old. North of College Point, on the eastern side of College Fiord, there is thick mature forest at sea level, as far as Photo Station K (Fig. 43). North of this, to a point opposite Smith Glacier, there are large groves, also at sea level, interspersed with open spaces; and a few scattered spruces grow up to the very edge of Harvard Glacier. Throughout Harvard Arm and lower College Fiord, the open spaces are covered by dense thickets of good-sized alders and willows and, except in the narrow barren zones of the cascading glaciers, there is no considerable area with as sparse growth as on the eastern side of Yale Arm. Thus Harvard Arm has much more forest than Yale Arm, and the eastern side is more thickly forested than the western where the cascading glaciers have broken it up. The conditions in Harvard Arm show clearly that Harvard Glacier has not recently extended southward beyond Pt. K., where the thick forest ends, and that for at least a century no advance has gone as far as the one which was in progress in 1910, contrasting, therefore, with the recent behavior of Yale Glacier. REASON FOB ADVANCES OF GLACIERS It is not perfectly clear why the glaciers of Unakwik Inlet and College Fiord are oscillating, some advancing and some retreating. The fundamental cause is, of course, involved with snow supply and with rate of melting as determined by temperature. Climatic records are so incomplete that we cannot settle this question and it is quite possible that, as in Yakutat Bay, the snow supply of any glacier, or group of ice tongues, might have been increased by avalanching during earthquakes, rather than by increase of precipitation. These glaciers of northern Prince William Sound lie entirely outside the zone affected by great avalanching during the Yakutat Bay earthquakes of September, 1899; but there have been a number of other severe, recent earthquakes in Alaska, whose origins were in or near the region of these advancing glaciers. One such, in October, 1900, was of considerable intensity. The Yakutat Bay glaciers which have so far advanced as a result of the 1899 earthquakes are of variable sizes and have responded in order of size, the smaller ones first.